To determine the effect of an evergreen windbreak on residential heat losses attributable to air infiltration, 8-m tall pine trees were arranged as an experimental windbreak to shelter a townhouse for 9 weeks. Air infiltration was measured continuously with SF6 as a tracer to compare air infiltration rates before and after the windbreak. For ASHRAE design wind and temperature conditions, air infiltration was reduced by 42%. As air infiltration accounts for about one-third of the total winter heat loss in these houses, a space heating reduction of about 14% is predicted under these conditions with the wind perpendicular to the windbreak. A dimensionless parameter was derived to distinguish between wind- and temperature-produced air infiltration and to determine the effect of wind direction on air infiltration:

M=V²/g*l*beta*deltaT

The parameter is the ratio of wind inertia to buoyancy effects: where V is wind speed, g the gravitational constant; lower-case "l", a characteristic length; beta, the coefficient of the thermal expansion of air; and deltaT, the indoor-outdoor temperature difference.